We applied CFD and AI‑enabled simulation to optimise a Y‑branch pipe that reduces head loss and mitigates waste deposition. The baseline geometry and several parametric variants (branch angle, blend radius, crotch curvature and leg offsets) were assessed across realistic flow rates and fluid properties. Loss mechanisms, separation and secondary flows were mapped to understand how local features drive pressure drop and low‑shear pockets where solids can accumulate. We built a fast‑running surrogate from the CFD design‑space to guide the search towards high‑performing regions using multi‑objective criteria, minimising loss coefficient while suppressing recirculation volume and wall‑shear deficits. Mesh/solver settings were tuned for reliable turnaround; verification checks ensured comparative trends were stable across the option set. Outputs included loss‑coefficient breakdowns, velocity/wall‑shear maps and deposition‑risk indicators aligned to manufacturing constraints. The selected concept offered a clear, manufacturable path to reduced energy consumption and improved cleanliness, with implementation guidance and a reproducible workflow for future iterations or scale changes.